SVM: Process data with identical fields-of-view

spacetelescope / drizzlepac

AstroDrizzle for HST images.

https://drizzlepac.readthedocs.io

BSD 3-Clause "New" or "Revised" License

50 stars 39 forks source link

SVM: Process data with identical fields-of-view #908

Closed stscijgbot-hstdp closed 2 years ago

stscijgbot-hstdp commented 3 years ago

Issue HLA-473 was created on JIRA by Warren Hack:

As reported by Tom Bair on 7-Jan-2021:

bq. ACS_969_01 failed this morning in HAP_SVM. I filed a JIRA ticket DPASMET-203 to document this event.

Discussing the cause of this Exception with Mihai, it turns out the sky matching code has a known and apparently insoluble failure mode triggered by this dataset related to exactly overlapping exposures. The best available solution would be to update Drizzlepac itself to look for and catch this specific Exception and change the sky matching mode.

stscijgbot-hstdp commented 3 years ago

Comment by Warren Hack on JIRA:

A new pytest needs to be added to provide test coverage for this error case. A suitable test would include data which has CR-SPLIT or similarly exactly overlapping observations. Ideally, new tests should be added for both standard pipeline processing as well as SVM mode testing, however, the minimum for this ticket would be to just cover the case that triggered this failure in pipeline processing (see DPASMET-203).

stscijgbot-hstdp commented 3 years ago

Comment by Warren Hack on JIRA:

Drizzlepac PR #⁠1002 was merged on 08/Apr/2021 to iterate over the sky subtraction modes in the event that the sky matching algorithm triggered an Exception these types of datasets.

stsci-hack commented 2 years ago

Resolved with #1002.